Sept., 1950 ship is obtained. Data for the branched chain isomers, however, indicate that the relative yields of hydrogen and methane are not proportional to the fractional concentration of C-H and C-CH3 bonds in the bombarded molecule. This selective decomposition of organic compounds by ionizing radiation has been recently discusseds in terms of competing free radical and rearrangement processes. To obtain information on the effect of isomerism on the radiation chemistry of organic acids, we have studied the decomposition of several series of isomeric acids in the ozonizer discharge which has been showns to produce chemical change by processes similar to those obtained with high-energy radiation. A previous paper' reported the decomposition of the toluic acids. The present paper reports a similar study of the valeric acid isomers. Gaseous decomposition products obtained from the valeric acids in the liquid state at 0' are shown in Table I. By comparing the CH4/ Hz yield ratios for the four isomeric acids, it is seen that the relative yield of methane to hydrogen is not proportional to the C-CHa/C-H bond ratio in the bombarded molecule. In fact, the methane to hydrogen yield ratio for trimethylacetic acid is actually smaller than the corresponding values observed with the other three compounds. The yield of methane relative to carbon dioxide plus carbon dioxide plus carbon monoxide also shows a minimum value with the tertiary acid. Increasing substitution on the alpha carbon atom results in a decrease in the probability of methane formation with a corresponding increase in the relative yield of higher hydrocarbons. A quantitative analysis of the higher hydrocarbon fraction was not undertaken; qualitative fractionation, however, indicated the presence of ethane and hydrocarbons with higher boiling points. A more detailed discussion of these reactions will be presented in a later paper.
NOTES
4303
Co., were used without further purification. n-Valeric acid (b. p. 186.2') was prepared by hydrolysis of n-valeronitrile and purified by fractional distillation. Trimethylacetic acid (b. p. 173-174') was prepared by the method of Puntambeker and Zoellner.* The all-glass discharge chamber,@the temperature of which was controlled by an ice-water-bath, was connected through stopcocks to a manifold, to a product gas storage bulb, and t o a Toepler pump which was used to transfer the decomposition products to a gas microanalyzer.10 The entire system could be evacuated to mm. by a mercury diffusion pump attached t o the manifold through a liquid-air trap. The transformer and experimental procedure were identical with those previously described.* (8) "Organic Syntheses," 2nd ed., Vol. I , John Wiley and Sons, Inc., New York, 1944,p. 524. (9) E. Warburg, 2.tech. Physik, 4, 450 (1923). (10) C. N. Stover, W. S. Partridge and W. M. Garrison, A n d . Chcm.. 01, 1013 (1949).
DEPARTMENT OF CHEMISTRY UNIVERSITY O F WYOMING LARAMIE, WYOMING RECEIVED MARCH17, 1950
Polymorphism of 17-Ethinylestradiol BY RICHARD PHEASANT
17-Ethinylestradiol has been .described with a melting point of 146O.l A second polymorphic form, m. p. 183' cor., has been encountered and conditions have been found which effect the interconversion of the two forms. Samples of pharmaceutical grade ethinylestradiol which melted in the normal manner a t 146' were sometimes observed to be slightly cloudy in appearance. Sustained heating a t 150 to 160' caused these samples to resolidify completely to acicular crystals. This crystalline form underwent no physical change on cooling. The melting point of this form was about 183', a t which temperature it melted completely to a clear liquid, which did not crystallize on cooling but formed a glassy mass. However, it could be made to crystallize as either the 146 or the 183' melting form by seeding with the desired form. TABLE I Seeding with a mixture of both forms, even a t a GASEOUSPRODUCTS FROM THE VALERIC ACIDS IN THE temperature below 140°, produced only the higher OZONIZER DISCHARGE (VOLUME %> melting form. MethylTriA methanol solution prepared from the highethylmethyln-Valeric Isovaleric acetic acetic melting form gave the low-melting form upon HnO 11.5 11.3 10.7 11.2 9.3 8.3 14.5 14.8 concentration and cooling. The mother liquor coz 10.0 10.0 10.7 9.8 10.2 10.7 18.3 19.3 co 16.6 15.3 22.0 22.4 26.8 25.5 22.6 22.6 was diluted with ether and upon standing for Hn . 54.3 54.6 40.7 40.6 38.0 40.1 17.7 17.1 several days gave crystals of the high-melting CHI 8.8 9.0 8.2 8.4 8.3 7.8 1.8 0.5 form. Other" hydroA sample of the 183' melting form gave the carbons 0.2 0.2 7.3 7.0 6.8 7.0 26.3 25.0 following analysis: Calcd. for CaoH~0a:C, 81.04; a Condensable in liquid air after removing water and H, 8.16. Found: C, 80.80; H, 8.30. Infrared carbon dioxide. absorption spectra were determined on samples Experimental of both forms, both in chloroform solution and in Isovaleric acid (b. p. 176176.5') and methylethylacetic acid (b. p. 173-174'), obtained from Eastman Kodak mineral oil suspension.2 The chloroform solutions gave identical absorption curves between 2 (6) M. Burton, J . Pkys. Chcr., 61, 786 (1847). and 12 microns. Curves from the mineral oil (6) G. Glockler and S. C. Lind, "The Blectrochemiitry of Gasand Other Dielectrics," John Wiley and Sono, Iac., New York, N. Y.,
lQ80. (7) C. N. Stover and W. M. Garrison, THISJOURNAL, 79, 2793 (1950).
(1) Inhoffen, Logemann, Hohlweg and Serini, Ber., T l , 1024 (1958). (2) The author i. indebted to W.
B. Tarplep, Research Labora. tories of this company, for the absorption spectra.
4304
~ 7 ~ 73 1 .
NOTES
suspensions were dissimilar only in the region from 8 to 12 microns, possibly due to the differences in crystal structure. Biological assays of the two forms showed no significant difference in estrogenic activity.3
bromide in the presence of oxygen was irradiated for 0.5-1 hour the tube was filled with brown vapors, which became more deeply colored as the time of irradiation increased. The mechanism by which the oxidation of aluminum bromide proceeds was not studied. It is very likely that the promoting ( 3 ) Modified Kahnt-Doisy method, by S. Margohn and M T. effect of the oxygen is due to the oxidation of Spoerlein, Biological Laboratories of this Company aluminum bromide with formation of bromine. QUALITYCONTROL DEPARTMEXT SCIIERINC CORPORATION The latter then reacts with methylcyclopentane BLOOMFIELD, N. J. RECEIVED MAY9, 1950 to form bromomethylcyclopentaane, which is a chain initiator for the isomerizati~n.~The oxidation of aluminum bromide in solution in hydroIsomerization of Saturated Hydrocarbons. carbons seems to proceed even in diffused light. VII1.I The Effect of Oxygen and The inhibiting effect of benzene upon the isomthe Isomerization of Methylcyclopen erization of methylcyclopentane is not entirely Presence of Aluminum Bromide unexpected; it is most likely due to the removal of BY HERMAN PINES,EUGENE ARISTOFF~ AND V. N. IPATIERF the chain initiator through the reaction with benzene, as has been shown previously.'~5 The promoting effect of oxygen upon the isomTABLE I erization of n-butane and n-pentane in the ISOMERIZATION OF METHYLCYCLOPENTANE presence of either aluminum chloride or aluminum Methylbromide has been reported previously.5 This cycloReactant charged Cyclopentane moles per 100 moles hexane study has now been extended to determine Irradi- Reac- used, of methylcyclopentane prowhether oxygen in the presence of aluminum ation, tion moles duced, Expt hours tube6 X 102 AlBrr Oz CSHs yo bromide but in .the absence of added hydrogen 1 18" P 2 22 2 01 0.18 0 0 bromide promotes the isomerization of saturated 2 18" P 3 10 4 00 1.05 0 15 cyclic hydrocarbons such as methylcyclopentane 3 18 P 2 11 2 . 0 2 0 18 0 6 to cyclohexane. The experiments were con4 IS Q 1.37 199 17 0 13 ducted in either quartz or Pyrex reaction tubes. 5 18 Q 1.04 4 02 .96 0 42 It was noticed that in diffused light, and a 6 I8 Q 2.60 2 04 .20 0.037 0 Pyrex reaction tube, with methylcyclopentane, 7 18 Q 1 00 3.98 .96 0.039 6 aluminum bromide and oxygen in a molal ratio a In experiments 1 and 2 the reaction tubes were not of 100:2 :0.2, isomerization of methylcyclopentane I, P, Pyrex reaction tube; Q, quartz reaction to cyclohexane does not occur. When the re- irradiated. tubc . action tube was exposed for eighteen hours to a Experimental quartz cadmium-mercury arc lamp 6y0 of the The high vacuum apparatus and procedure have been methylcyclopentanewas isomerized; the extent of described previously.1,6 Linde oxygen was introduced into isomerization was increased to 13% when a quartz the apparatus through a phosphorus pentoxide drying tube. After being measured in the calibrated portion of reaction tube instead of a Pyrex one was used. the Topler pump, the oxygen was transferred to an amThe introduction of a larger amount of oxygen poule having a thin walled break-off. The sealed aminto a Pyrex reaction zone, namely, 1 mole equiv- poule was carefully placed in the reaction tube, then the alent of oxygen per 100 moles of methylcyclo- latter evacuated. After the other reactants had been the reaction tube was sealed off then shaken, in pentane and 4 moles of aluminum bromide, caused added, order to break the thin bulb of the ampoule containing the isomerization of 15% of methylcyclopentane; oxygen. The composition of the hydrocarbons obtained when a quartz reaction tube was used and irradi- from the reaction was determined by means of infrared absorption spectra. ated, 42% of cyclohexane was formed. In line with previous 0bservations1*~~5 it was (6) H. Pines, B. M Abraham and V. N. Ipatieff, ibid., 70, 1742 noticed that the presence of about 0.04 mole per (1948). cent. of benzene in methykyclapentane greatly THEIPATIEPFHXG3 PRESSUREAND LABORATORY reduces the degree of isomerization. The experi- CATALYTIC DEPARTMENT OF CHEMISTRY mental results are summarized in Table I. UNIVERSITY During the course of this study it was observed NORTHWESTERN EVANSTON, ILLINOIS RECEIVED MARCH 22, 1950 that aluminum bromide per se placed in a quartz tube and irradiated did not change coloration. It was noticed however that when aluminum The Structure of Methyl Acetylacrylate] (1) For paper V I 1 of this series see H. Pines, E. Aristoffand V. N. Ipatieff. THIS JOURNAL, 78,4055 (1950). (2) Universal Oil Products Company Predoctoral Research Fellow 1947-1949. (3) H Pines and R. C. Wackher, Tzus JOURNAL, 66,599 (1946). (4) J. M . Mavrty, H. Pines, R. C. Wackher and J. A. Brooks, I d . En;. Chcm., 10, 2374 (1948). (5) E. Pines, E. Ariatoff m d V. ET. Xpaficff, T a a JOWRNAL, T I , 749 (19491.
BY SAMUEL RAYMOND
Methyl acetylacrylate was first prepared in 1914 by Pauly, Gilmour and Willla by the dehydro(1) This work waa supported by a grant from the National Institutes of Hedth and from the John and Mary R. Mar!& Poundation. 1.1( Pfial~,Wmour nnd Will, Ann., 408,1x9 (1814).
